A mere correspondence between the electron statistics and the photon one vanishes in the feedback loop (FBL). It means that the direct photodetection, supplying us with the electron statistics, does not provide us with a wished information about the laser photon statistics. For getting this information we should think up another measurement procedure, and we in the article suggest applying the three-level laser as a auxiliary measuring device. This laser has impressive property, namely, its photon statistics survive information about the initial photon statistics of the laser which excites coherently the three-level medium. Thus, if we choose the laser in the FBL as exciting the three-level laser, then we have an possibility to evaluate its initial photon statistics by means of direct detecting the three-level laser emission. Finally, this approach allows us to conclude the feedback is not capable of creating a regularity in the laser light beam. Contrary, the final photon fluctuations turn out to be always even bigger. The mentioned above feature of the three-level laser takes place only for the strong interaction between the lasers (exciting and excited). It means the initial state of the exciting laser is changed dramatically, so our measurement procedure can not be identified with some non-demolition one.